TUBECLINIC

LINZ / AUSTRIA

substituting a choke
By Barbara E. Gerhold, TUBECLINIC
I again wanted to thank Mr. van Hall for his splendid idea to feed the pre-stages (B1+) from
the center point of the two serial caps of the B+ supply!
but there was a little problem with increased hum at the output! Because my efficient
horn speakers reflected this problem in a very audible way, I turned back to the connection
from before, taking B1+ from the B+ supply via a heavy resistor chain.
Knowing that this special idea could be brought to a better result, I started to research,
which design could give a better hum and noise figure. The found solution is again a
gyrator, which might be usable also for other applications, so I started to write a new
article about.
This B1+ gyrator circuit is intended to replace the choke 10H/25mA/EI55 from the circuit
shown here.
Have a look at the schematic:

V1.0 2013 by Barbara E. Gerhold, TUBECLINIC, Linz/Austria; mailto: support@tubeclinic.com

Page 1

TUBECLINIC
LINZ / AUSTRIA
Function:
Because the above schematic consists of two identical circuits, I shall explain the circuitry
only by the one for the left channel.
Substituting the choke, the terminals JP1 and JP2 are connected instead of the choke while
JP3 is connected to GND.
The supply comes from the center point of the two caps in series and it measures half of
the B+, in this case is half of the +930VDC -> +465VDC. The current across the resistor chain
R2, R4, R5, R6, R7 (1W, 1%, metal film) should be set to app. 1mA, so the total dissipation
within will stay to less than W. Values can be found in a table in behind. Because this
design was intended to be versatile, you can download a little EXCEL spreadsheet here,
ready to calculate the needed values for your special application.
The voltage across R2 should be kept to at least 5 6VDC, so also the voltage across the
power MOSFET Q1 will be app. 10VDC. The resistor R3 (0,5W, 1%, metal film) charges the
cap C3. It can be of large value, since the gate of the MOSFET is extremely high resistant. It
resembles only a gate-charge of some pF. Resistor R1 (0,5W, 1%, metal film) shall only
prevent oscillation.
Cap C2 is used to filter noise from the incoming supply voltage. It widely improves the
noise figure. C1 in parallel to C3 serves the same reason. Both should be FKP-type.
Power-up delay of this circuitry will be app. 10 seconds. This will also prevent bumps,
coming from the charging of coupling caps.
The used MOSFET (IRFP450, datasheet here) will stand 500VDC across drain/source. If you
wanted to use this circuit for a bigger voltage, you should consider a type showing bigger
Uds. Furthermore, the MOSFET has to be mounted (insulated!) to a suitable heat sink (i.e.
the amp-chassis itself). In the case of MA845, the power dissipation will be app. 0,2W
during idling operation, but there will be some more during switch-on and if the circuit
had to regulate, i.e. due to mains variations, it also will be more. If you increased the value
of R2, the MOSFET will drop more voltage, so you could also adjust a personal B1+ within
your amp. This again will lead to bigger dissipation and will increase heat. So be aware to
apply a proper heat sink!
The basic function of any gyrator is that an inverting, amplifying part (MOSFET, transistor)
with a capacitance at the gate (base), acts like an inductance across drain/source
(collector/emitter). The value of inductance, which is achievable by this configuration, is
much bigger than any iron-choke could become. So the hum and noise reduction is
outstanding.
As you already know all information about the printable, full-size schematic and the PCB
is again situated in a binder enclosed to this website.
If you have any further question, you can contact me via support@tubeclinic.com .

V1.0 2013 by Barbara E. Gerhold, TUBECLINIC, Linz/Austria; mailto: support@tubeclinic.com

Page 2

TUBECLINIC
LINZ / AUSTRIA
TABLE
B1+ GYRATOR; voltage-drop calculator, power-dissipation
U_in

U_out

R2

R4,R5,R6,R7

drop

current

[VDC]

[VDC]

[kohm]

[kohm]

[VDC]

[mA]

250
300
350
400
460
460

239,6
289,7
339,9
389,0
449,3
433,0

6,8
6,8
6,8
8,2
6,8
22

56
68
82
100
100
100

10,4
10,3
10,1
11,0
10,7
27,0

30
50
20
20
18
18

Power in [mW]
resistors MOSFET
207,7
697,0
133,9
163,3
131,8
136,7

311,0
515,9
202,2
220,7
192,4
485,7

TOTAL
518,7
1212,9
336,1
384,0
324,2
622,4

2013 Barbara E. Gerhold; TUBECLINIC; Linz/Austria

The EXCEL spreadsheet comes here!

V1.0 2013 by Barbara E. Gerhold, TUBECLINIC, Linz/Austria; mailto: support@tubeclinic.com

Page 3